Re: [RFC] Should we create a raw input interface for IR's ? - Was: Re: [PATCH 1/3 v2] lirc core device driver infrastructure

[Dmitry Torokhov <dmitry.torokhov@gmail.com>]

On Thu, Nov 26, 2009 at 10:08:29PM -0500, Jon Smirl wrote:
> On Thu, Nov 26, 2009 at 9:28 PM, Jarod Wilson <jarod@wilsonet.com> wrote:
> >> No, at present we expect 1:1 button->event mapping leaving macro
> >> expansion (i.e. KEY_PROG1 ->  "do some multi-step sequence" to
> >> userspace).
> >
> > Hm. So ctrl-x, alt-tab, etc. would have to be faked in userspace somehow.
> > Bummer.
> 
> That is scripting. Scripting always needs to be done in user space.
> 
> In the code I posted there is one evdev device for each configured
> remote. Mapped single keycodes are presented on these devices for each
> IR burst. There is no device for the IR receiver.  A LIRC type process
> could watch these devices and then execute scripts based on the
> keycodes reported.
> 
> The configfs model is very flexible. You could make a "remote" that
> translates the UP/DOWN buttons of several different remotes into
> KEY_UP/DOWN.  That lets several different remotes control the same
> app.
> 
> Sure it is clunky to play with IR hex codes and keycodes in the
> configfs mapping dir. If you don't like it write a GUI app for
> manipulating the codes. GUI would then generate a script for udev to
> run which builds the configfs entries.
> 
> Maybe I should rename those directory entries to "app" instead of
> "remote". They contain the mappings from IR hex codes to keycodes that
> an app is interested in. Usually there is a 1:1 correspondence between
> remote and app but there doesn't have to be.
> 

Maybe we should revisit Jon's patchset as well. Regretfully I did not
have time to do that when it was submitted the last time.

-- 
Dmitry